Can Mentoring Help Female Assistant Professors? Interim Results from a Randomized Trial

Mentoring young female professors in economics, an academic field in which women are historically underrepresented, can lead to an increase in top-tier publications and grants for these women relative to their peers without mentors.

FindingsMethodology

Women are significantly underrepresented in tenured positions in economics. Previous studies have found a “leaky pipeline” for women in the field. For example, one study finds  a substantial gender gap in promotion to tenure  of 21 percentage points. Additionally, women  are significantly less likely to be promoted in economics relative to other fields such as  political science, statistics, life science, physical science and engineering.

One strategy to address the gender gap in promotion to tenure is to mentor new female faculty. This paper presents results of a randomized evaluation of CeMENT workshops. These workshops were designed to expose participants to role models by placing them in small group sessions with senior female economists and to give them the opportunity to get feedback on their work. Women applied to participate in the workshop and acceptance to the workshop as random. This study investigates the effectiveness of CeMENT in advancing women’s academic careers in economics.

Findings

Overall, the CeMENT mentoring workshops for female economics professors increased the number of top-tier publications and grants for women involved. The authors use data from three cohorts of women who participated in the workshop in 2004, 2006 and 2008. Comparing women who participated in the workshop to women who did not participate, the authors measure the effect of workshop participation on number of grants, any top tier publications, and total number of publications. The authors find that:

  • One year after the intervention: There is no effect on the outcomes of interest, which is not surprising given the timelines for grant writing and publications.
  • Three years after the intervention: Workshop participants were 20 percentage points more likely to have a top-tier publication, an average of two more publications, and a positive effect on the number of grants, than those who hadn’t been mentored.
  • Five years after the intervention: Workshop participants had significant gains in grants and publications including .4 more NSF or NIH grants. Participants were 25 percentage points more likely to have any top-tier publication than their peers. Participants also had an average of three more overall publications than their peers in the control group.
  • These effects marginally diminished when the authors controlled for the pre-treatment characteristics of participants including whether they had a job at a top-ten department and their number of top-tier publications prior to participating in the workshop.
  • However, controlling for pre-treatment characteristics, workshop participants were 20 percentage points more likely to have a top-tier publication and have 2.7 more publications overall, fiver years after the workshop, compared to controls, but there were no longer any significant effects on grants.

In this study, the authors present data from three cohorts: 2004, 2006 and 2008. They have interim results after one year for all three workshops, after three years for the cohorts from 2004 and 2006 and after five years for 2004 cohort. Given that there is only five-year data from one cohort, it is still too early to say whether the intervention will have a significant effect on either the probability that women stay in academia or the probability that they receive tenure. However, the early positive results around publications and grants suggest that participants are on a trajectory towards increased professional success.

Methodology

Individuals apply to participate in the CeMENT workshops. Applicants were divided into groups by research subfield. Within each research subfield, applicants were assigned to either “treatment” – those who got to participate in the mentoring sessions at the American Economic Association’s annual meeting –  or “control” – those who did not get accepted to participate. They assigned more than half of the participants to the treatment group because they wanted to maximize access to the treatment. All participants were told there was excess demand to participate in the program and that participants were randomly chosen from the applicant pool. Except for cohort two, which had a disproportionate amount of academics who had already published in top-tier publications and who were from top ten departments, the control and treatment groups had no statistically significant differences for a variety of characteristics.

Participants circulated a research paper or other related work, such as a grant proposal, to their groups before the workshop. During the workshop, the small groups met to discuss and provide feedback on each participant’s work. This averaged an hour per participant. There were also plenary sessions consisting of panels of the senior mentors during the workshop. Topics included research and publishing, getting grants, professional exposure, teaching, the tenure process and work-life balance.

Participants rated their experience highly after participating in the small groups and plenaries (average score of 6.63/7). To gauge the effects of the treatment, the authors gathered data on the participants and control group members one, three and five years afterward the workshops.


MLA: Blau, Francine D. et al. “Can Mentoring Help Female Assistant Professors? Interim Results from a Randomized Trial.” American Economic Review 100.2 (2010): 348–52. AEA Web. Web.
APA: Blau, F. D., Currie, J. M., Croson, R. T. A., & Ginther, D. K. (2010). Can Mentoring Help Female Assistant Professors? Interim Results from a Randomized Trial. American Economic Review, 100(2), 348–52. http://doi.org/10.1257/aer.100.2.348
Chicago: 1. Francine D. Blau et al., “Can Mentoring Help Female Assistant Professors? Interim Results from a Randomized Trial,” American Economic Review 100, no. 2 (2010): 348–52, doi:10.1257/aer.100.2.348.